PUST: Transferring Update Signals via Proxy Models

KnowledgeXLab · hf · 2026-07-14

This paper introduces a novel post-training paradigm, Proxy-guided Update Signal Transfer (PUST), aiming to decouple "exploration" from "distribution alignment".

Core Idea

Main Benefits

Experimental Conclusions

The authors reframe post-training from a "monolithic online optimization" process into a more modular, reusable, and cost-effective pipeline.

Related event: PUST: New Paradigm Uses Small Models to Guide Large Model Training(2 posts)→

Original post →

More from Research

Research channel →